AX1200 - REVO 85+ 1250 - or???

[_fabry_]

hi guys..
i'm new there and i'm writing from Italy... (sry for my english...)

i've a stressing doubt that stress my mind...

i wanna change my glorious HX1000, cause my system under heavy OC (cpu vga & ram) needs a ton of current and my PSU don't seems to be so good for that performance level... some reboot and freeze issues told me that the psu may be the problem...

i've red all the (wonderful and complete) reviews of the PSU made by JONNY but i'm still doubtful about purchasing...

looking at my system and my OC attitude and leaving apart the price-aestethics-noise level-fan boy-brand questions and talking about pure prestations, which one you would choose:

CORSAIR AX1200
ENERMAX REVOLUTION 85+ 1250W


thanks a lot!!

[-The_Mask-]

Reboot with or without a BSOD?

[_fabry_]

without... a curt reboot

[jonnyGURU]

Just to be sure.....

Under "Advanced System Settings", under "System Failure", the box for "Automatically Restart" is NOT checked?

Just want to make sure because PSU overload or failure does not typically result in a reboot or hard lock. Usually an overloaded PSU will just shut down. It shouldn't reboot or even power back up again unless the I/O switch on the back of the unit is toggled after such a shut down.

[_fabry_]

hey, hi Jonny, (what a honour!! )

yeah my OS is ok, checked and re-checked out... it's absolutely an hardware issue...
if you have a couple of minutes i'll explain it...
as you (obviously ) know, the hx1000 is simply 2x500w psu, so we have two different 12v rail (one for every 2 of 500) to power the vgas and the cpu...
using 2 current-hungry gtx580 OCed, i had to subdivide the two cards for each of the two rails so the first 580 is on 12v1 and the 2nd 580 is on the 12v2... cuz surely (and tested by me) the 2 gtx can't be feed by a single rail (such as 12v2,as corsair raccomands),500w are not enough..
so the 12v2 is used only for the vga but 12v1 is used for the vga and the cpu... both very OCed...
here comes the problems....
using hwinfo that helps me monitoring many values about my pc under stress i noticed that the 12v1 rail under gaming (crysis) has a major vdrop than the 12v2...
so,increasing the OC, when under heavy load the 12v1 goes under the 11.9x volts the system crash or freeze or some...

my first impression is simply "my hx1000 have been maxed out!"...
now my doubt is: my issue is generated by the 580s oc wall or by my hx1000 that cannot supply all the current the system needs??
how can i determinate it??
may i have an advice by you??

thanx guys!!

[jonnyGURU]

I wasn't assuming there was a problem with your OS, OS installation, etc.

By default, Windows from XP on up is set to reboot if there is any error. You have to manually go in and tell it NOT to reboot and to give you a BSOD instead.

So I was asking if that box was unchecked so we can see if the machine reboots or gives you a BSOD.

Yes, the HX1000 is "technically" two 500W PSU's wedged together, but that doesn't change what the expected behavior should be if one of the rails is overloaded. It should shut down. Rebooting or the PC locking up is a not a normal symptom.

As for your voltage drops: You're getting those from HWINFO which, in turn, gets that info from the BIOS. That doesn't convince me of anything.

[allikat]

The theoretical power draw from that system should be under 800w, plus a bit for the heavy overclock, and you shouldn't be over-stressing that HX1000. I'd suggest that the overclocks are a bit high on the GPUs to be in SLi mode, they will likely be 100% stable on a single card basis, but SLi puts a bigger strain on the primary card, and not usually one that can be fixed with more voltage either.
Back the clocks off a touch and your problems should be significantly less.

[_fabry_]

hi guys
and thank you for the replies!!

so, in your opinion, the psu is not the real problem??
the crashes and freeezes is caused by the vga's oc limit??
i've not to look at the voltages on the 12 detected in windows by the utilities??
a single 12v rail of 500w power may supply a gtx580 + 2500k strongly oced??

how may i test this??

thanks guys, thanks a lot!!

[Original Sin]

Originally Posted by _fabry_

hi guys
and thank you for the replies!!

so, in your opinion, the psu is not the real problem??
the crashes and freeezes is caused by the vga's oc limit??
i've not to look at the voltages on the 12 detected in windows by the utilities??
a single 12v rail of 500w power may supply a gtx580 + 2500k strongly oced??

how may i test this??

thanks guys, thanks a lot!!

You're looking at this the wrong way, this is a unit with no OCP, and OPP set somewhere above 120A(iirc)... in other words this unit can supply far more than it's labeled power, so much so that you'll never run into any sort or overload protection with your current hardware... I doubt it's a regulation issue either, since pretty much all the PUC units we've seen keep the voltages in spec even at 1.5KW.
I'd suggest you look for the problem elsewhere and not the PSU... that being said, a newer, more efficient unit wouldn't hurt.

[_fabry_]

hi and thank you for the reply...


you (all) are helping me very much an i'm very grateful.

with your advices in my mind, i've grabbed my 12v tester in hand and made some ... tests...

in full load i've red the value at the 6 pin pci-ex supply cable from the supply unit and it gave me 12,07 costant... bah

so i've placed it on the soldered eight pin of the vga's pcb (12v1) and it tells me that on max stress the lower voltage was 11.95v.

are my tests reliable or are there one more effective and certained (but affordable too ) method to measure the 12v that supply my PSU under load??

thanks a ton!!